Template fix for broken publisher links
--- a/README.rst
+++ b/README.rst
@@ -32,6 +32,7 @@
googleanalytics.id = UA-1010101-1
googleanalytics.account = Account name (e.g. data.gov.uk, see top level item at https://www.google.com/analytics)
+ googleanalytics.token.filepath = ~/pyenv/token.dat
ga-report.period = monthly
ga-report.bounce_url = /
@@ -82,13 +83,17 @@
$ paster getauthtoken --config=../ckan/development.ini
+Now ensure you reference the correct path to your token.dat in your CKAN config file (e.g. development.ini)::
+
+ googleanalytics.token.filepath = ~/pyenv/token.dat
+
Tutorial
--------
Download some GA data and store it in CKAN's database. (Ensure your CKAN pyenv is still activated, run the command from ``src/ckanext-ga-report``, alter the ``--config`` option to point to your site config file) and specifying the name of your auth file (token.dat by default) from the previous step::
- $ paster loadanalytics token.dat latest --config=../ckan/development.ini
+ $ paster loadanalytics latest --config=../ckan/development.ini
The value after the token file is how much data you want to retrieve, this can be
--- a/ckanext/ga_report/command.py
+++ b/ckanext/ga_report/command.py
@@ -1,5 +1,8 @@
import logging
import datetime
+import os
+
+from pylons import config
from ckan.lib.cli import CkanCommand
# No other CKAN imports allowed until _load_config is run,
@@ -58,20 +61,17 @@
"""Get data from Google Analytics API and save it
in the ga_model
- Usage: paster loadanalytics <tokenfile> <time-period>
+ Usage: paster loadanalytics <time-period>
- Where <tokenfile> is the name of the auth token file from
- the getauthtoken step.
-
- And where <time-period> is:
+ Where <time-period> is:
all - data for all time
latest - (default) just the 'latest' data
YYYY-MM - just data for the specific month
"""
summary = __doc__.split('\n')[0]
usage = __doc__
- max_args = 2
- min_args = 1
+ max_args = 1
+ min_args = 0
def __init__(self, name):
super(LoadAnalytics, self).__init__(name)
@@ -92,19 +92,25 @@
from download_analytics import DownloadAnalytics
from ga_auth import (init_service, get_profile_id)
+ ga_token_filepath = os.path.expanduser(config.get('googleanalytics.token.filepath', ''))
+ if not ga_token_filepath:
+ print 'ERROR: In the CKAN config you need to specify the filepath of the ' \
+ 'Google Analytics token file under key: googleanalytics.token.filepath'
+ return
+
try:
- svc = init_service(self.args[0], None)
+ svc = init_service(ga_token_filepath, None)
except TypeError:
print ('Have you correctly run the getauthtoken task and '
- 'specified the correct token file?')
+ 'specified the correct token file in the CKAN config under '
+ '"googleanalytics.token.filepath"?')
return
downloader = DownloadAnalytics(svc, profile_id=get_profile_id(svc),
delete_first=self.options.delete_first,
skip_url_stats=self.options.skip_url_stats)
- time_period = self.args[1] if self.args and len(self.args) > 1 \
- else 'latest'
+ time_period = self.args[0] if self.args else 'latest'
if time_period == 'all':
downloader.all_()
elif time_period == 'latest':
--- a/ckanext/ga_report/controller.py
+++ b/ckanext/ga_report/controller.py
@@ -22,12 +22,29 @@
def _month_details(cls):
- '''Returns a list of all the month names'''
+ '''
+ Returns a list of all the periods for which we have data, unfortunately
+ knows too much about the type of the cls being passed as GA_Url has a
+ more complex query
+
+ This may need extending if we add a period_name to the stats
+ '''
months = []
- vals = model.Session.query(cls.period_name).filter(cls.period_name!='All').distinct().all()
+ day = None
+
+ vals = model.Session.query(cls.period_name,cls.period_complete_day)\
+ .filter(cls.period_name!='All').distinct(cls.period_name)\
+ .order_by("period_name desc").all()
+ if vals and vals[0][1]:
+ day = int(vals[0][1])
+ ordinal = 'th' if 11 <= day <= 13 \
+ else {1:'st',2:'nd',3:'rd'}.get(day % 10, 'th')
+ day = "{day}{ordinal}".format(day=day, ordinal=ordinal)
+
for m in vals:
months.append( (m[0], _get_month_name(m[0])))
- return sorted(months, key=operator.itemgetter(0), reverse=True)
+
+ return months, day
class GaReport(BaseController):
@@ -56,7 +73,7 @@
# Get the month details by fetching distinct values and determining the
# month names from the values.
- c.months = _month_details(GA_Stat)
+ c.months, c.day = _month_details(GA_Stat)
# Work out which month to show, based on query params of the first item
c.month_desc = 'all months'
@@ -220,7 +237,7 @@
# Get the month details by fetching distinct values and determining the
# month names from the values.
- c.months = _month_details(GA_Url)
+ c.months, c.day = _month_details(GA_Url)
# Work out which month to show, based on query params of the first item
c.month = request.params.get('month', '')
@@ -278,7 +295,7 @@
# Get the month details by fetching distinct values and determining the
# month names from the values.
- c.months = _month_details(GA_Url)
+ c.months, c.day = _month_details(GA_Url)
# Work out which month to show, based on query params of the first item
c.month = request.params.get('month', '')
--- a/ckanext/ga_report/download_analytics.py
+++ b/ckanext/ga_report/download_analytics.py
@@ -126,7 +126,7 @@
ga_model.update_publisher_stats(period_name) # about 30 seconds.
log.info('Downloading and storing analytics for site-wide stats')
- self.sitewide_stats( period_name )
+ self.sitewide_stats( period_name, period_complete_day )
log.info('Downloading and storing analytics for social networks')
self.update_social_info(period_name, start_date, end_date)
@@ -153,7 +153,8 @@
data = collections.defaultdict(list)
rows = results.get('rows',[])
for row in rows:
- data[_normalize_url(row[0])].append( (row[1], int(row[2]),) )
+ url = _normalize_url('http:/' + row[0])
+ data[url].append( (row[1], int(row[2]),) )
ga_model.update_social(period_name, data)
@@ -194,7 +195,7 @@
if 'url' in data:
ga_model.update_url_stats(period_name, period_complete_day, data['url'])
- def sitewide_stats(self, period_name):
+ def sitewide_stats(self, period_name, period_complete_day):
import calendar
year, month = period_name.split('-')
_, last_day_of_month = calendar.monthrange(int(year), int(month))
@@ -205,7 +206,7 @@
'_locale_stats', '_browser_stats', '_mobile_stats']
for f in funcs:
log.info('Downloading analytics for %s' % f.split('_')[1])
- getattr(self, f)(start_date, end_date, period_name)
+ getattr(self, f)(start_date, end_date, period_name, period_complete_day)
def _get_results(result_data, f):
data = {}
@@ -214,7 +215,7 @@
data[key] = data.get(key,0) + result[1]
return data
- def _totals_stats(self, start_date, end_date, period_name):
+ def _totals_stats(self, start_date, end_date, period_name, period_complete_day):
""" Fetches distinct totals, total pageviews etc """
results = self.service.data().ga().get(
ids='ga:' + self.profile_id,
@@ -224,7 +225,8 @@
max_results=10000,
end_date=end_date).execute()
result_data = results.get('rows')
- ga_model.update_sitewide_stats(period_name, "Totals", {'Total page views': result_data[0][0]})
+ ga_model.update_sitewide_stats(period_name, "Totals", {'Total page views': result_data[0][0]},
+ period_complete_day)
results = self.service.data().ga().get(
ids='ga:' + self.profile_id,
@@ -239,7 +241,7 @@
'New visits': result_data[0][2],
'Total visits': result_data[0][3],
}
- ga_model.update_sitewide_stats(period_name, "Totals", data)
+ ga_model.update_sitewide_stats(period_name, "Totals", data, period_complete_day)
# Bounces from / or another configurable page.
path = '/%s%s' % (config.get('googleanalytics.account'),
@@ -261,10 +263,11 @@
bounces, total = [float(x) for x in result_data[0][1:]]
pct = 100 * bounces/total
log.info('%d bounces from %d total == %s', bounces, total, pct)
- ga_model.update_sitewide_stats(period_name, "Totals", {'Bounce rate (home page)': pct})
-
-
- def _locale_stats(self, start_date, end_date, period_name):
+ ga_model.update_sitewide_stats(period_name, "Totals", {'Bounce rate (home page)': pct},
+ period_complete_day)
+
+
+ def _locale_stats(self, start_date, end_date, period_name, period_complete_day):
""" Fetches stats about language and country """
results = self.service.data().ga().get(
ids='ga:' + self.profile_id,
@@ -279,16 +282,16 @@
for result in result_data:
data[result[0]] = data.get(result[0], 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Languages", data)
+ ga_model.update_sitewide_stats(period_name, "Languages", data, period_complete_day)
data = {}
for result in result_data:
data[result[1]] = data.get(result[1], 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Country", data)
-
-
- def _social_stats(self, start_date, end_date, period_name):
+ ga_model.update_sitewide_stats(period_name, "Country", data, period_complete_day)
+
+
+ def _social_stats(self, start_date, end_date, period_name, period_complete_day):
""" Finds out which social sites people are referred from """
results = self.service.data().ga().get(
ids='ga:' + self.profile_id,
@@ -304,10 +307,10 @@
if not result[0] == '(not set)':
data[result[0]] = data.get(result[0], 0) + int(result[2])
self._filter_out_long_tail(data, 3)
- ga_model.update_sitewide_stats(period_name, "Social sources", data)
-
-
- def _os_stats(self, start_date, end_date, period_name):
+ ga_model.update_sitewide_stats(period_name, "Social sources", data, period_complete_day)
+
+
+ def _os_stats(self, start_date, end_date, period_name, period_complete_day):
""" Operating system stats """
results = self.service.data().ga().get(
ids='ga:' + self.profile_id,
@@ -322,17 +325,17 @@
for result in result_data:
data[result[0]] = data.get(result[0], 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Operating Systems", data)
+ ga_model.update_sitewide_stats(period_name, "Operating Systems", data, period_complete_day)
data = {}
for result in result_data:
if int(result[2]) >= MIN_VIEWS:
key = "%s %s" % (result[0],result[1])
data[key] = result[2]
- ga_model.update_sitewide_stats(period_name, "Operating Systems versions", data)
-
-
- def _browser_stats(self, start_date, end_date, period_name):
+ ga_model.update_sitewide_stats(period_name, "Operating Systems versions", data, period_complete_day)
+
+
+ def _browser_stats(self, start_date, end_date, period_name, period_complete_day):
""" Information about browsers and browser versions """
results = self.service.data().ga().get(
ids='ga:' + self.profile_id,
@@ -349,14 +352,14 @@
for result in result_data:
data[result[0]] = data.get(result[0], 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Browsers", data)
+ ga_model.update_sitewide_stats(period_name, "Browsers", data, period_complete_day)
data = {}
for result in result_data:
key = "%s %s" % (result[0], self._filter_browser_version(result[0], result[1]))
data[key] = data.get(key, 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Browser versions", data)
+ ga_model.update_sitewide_stats(period_name, "Browser versions", data, period_complete_day)
@classmethod
def _filter_browser_version(cls, browser, version_str):
@@ -380,7 +383,7 @@
ver = ver[0] + ver[1] + 'X' * num_hidden_digits
return ver
- def _mobile_stats(self, start_date, end_date, period_name):
+ def _mobile_stats(self, start_date, end_date, period_name, period_complete_day):
""" Info about mobile devices """
results = self.service.data().ga().get(
@@ -397,13 +400,13 @@
for result in result_data:
data[result[0]] = data.get(result[0], 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Mobile brands", data)
+ ga_model.update_sitewide_stats(period_name, "Mobile brands", data, period_complete_day)
data = {}
for result in result_data:
data[result[1]] = data.get(result[1], 0) + int(result[2])
self._filter_out_long_tail(data, MIN_VIEWS)
- ga_model.update_sitewide_stats(period_name, "Mobile devices", data)
+ ga_model.update_sitewide_stats(period_name, "Mobile devices", data, period_complete_day)
@classmethod
def _filter_out_long_tail(cls, data, threshold=10):
--- a/ckanext/ga_report/ga_model.py
+++ b/ckanext/ga_report/ga_model.py
@@ -9,6 +9,8 @@
import ckan.model as model
from ckan.lib.base import *
+
+log = __import__('logging').getLogger(__name__)
def make_uuid():
return unicode(uuid.uuid4())
@@ -45,6 +47,7 @@
Column('id', types.UnicodeText, primary_key=True,
default=make_uuid),
Column('period_name', types.UnicodeText),
+ Column('period_complete_day', types.UnicodeText),
Column('stat_name', types.UnicodeText),
Column('key', types.UnicodeText),
Column('value', types.UnicodeText), )
@@ -132,7 +135,7 @@
return None, publisher_match.groups()[0]
return None, None
-def update_sitewide_stats(period_name, stat_name, data):
+def update_sitewide_stats(period_name, stat_name, data, period_complete_day):
for k,v in data.iteritems():
item = model.Session.query(GA_Stat).\
filter(GA_Stat.period_name==period_name).\
@@ -142,11 +145,13 @@
item.period_name = period_name
item.key = k
item.value = v
+ item.period_complete_day = period_complete_day
model.Session.add(item)
else:
# create the row
values = {'id': make_uuid(),
'period_name': period_name,
+ 'period_complete_day': period_complete_day,
'key': k,
'value': v,
'stat_name': stat_name
@@ -212,7 +217,7 @@
'period_complete_day': 0,
'url': url,
'pageviews': sum([int(e.pageviews) for e in entries]) + old_pageviews,
- 'visits': sum([int(e.visits) for e in entries]) + old_visits,
+ 'visits': sum([int(e.visits or 0) for e in entries]) + old_visits,
'department_id': publisher,
'package_id': package
}
@@ -344,20 +349,33 @@
model.Session.commit()
def get_score_for_dataset(dataset_name):
+ '''
+ Returns a "current popularity" score for a dataset,
+ based on how many views it has had recently.
+ '''
import datetime
now = datetime.datetime.now()
- period_names = ['%s-%02d' % (now.year, now.month),
- '%s-%02d' % (now.year, now.month-1)]
-
- entry = model.Session.query(GA_Url)\
- .filter(GA_Url.period_name==period_names[0])\
- .filter(GA_Url.package_id==dataset_name).first()
- score = int(entry.pageviews) if entry else 0
-
- entry = model.Session.query(GA_Url)\
- .filter(GA_Url.period_name==period_names[1])\
- .filter(GA_Url.package_id==dataset_name).first()
- val = int(entry.pageviews) if entry else 0
- score += val/2 if val else 0
-
- return 0
+ last_month = now - datetime.timedelta(days=30)
+ period_names = ['%s-%02d' % (last_month.year, last_month.month),
+ '%s-%02d' % (now.year, now.month),
+ ]
+
+ score = 0
+ for period_name in period_names:
+ score /= 2 # previous periods are discounted by 50%
+ entry = model.Session.query(GA_Url)\
+ .filter(GA_Url.period_name==period_name)\
+ .filter(GA_Url.package_id==dataset_name).first()
+ # score
+ if entry:
+ views = float(entry.pageviews)
+ if entry.period_complete_day:
+ views_per_day = views / entry.period_complete_day
+ else:
+ views_per_day = views / 15 # guess
+ score += views_per_day
+
+ score = int(score * 100)
+ log.debug('Popularity %s: %s', score, dataset_name)
+ return score
+
--- a/ckanext/ga_report/helpers.py
+++ b/ckanext/ga_report/helpers.py
@@ -50,9 +50,12 @@
dataset = model.Package.get(ga_url.url[len('/dataset/'):])
if dataset and not dataset.state == 'active':
dataset = None
- count += 1
- if count > 10:
- break
+ # When testing, it is possible that top datasets are not available
+ # so only go round this loop a few times before falling back on
+ # a random dataset.
+ count += 1
+ if count > 10:
+ break
if not dataset:
# fallback
dataset = model.Session.query(model.Package)\
--- a/ckanext/ga_report/templates/ga_report/ga_util.html
+++ b/ckanext/ga_report/templates/ga_report/ga_util.html
@@ -5,6 +5,14 @@
xmlns:xi="http://www.w3.org/2001/XInclude"
py:strip=""
>
+
+<select name="month" py:def="month_selector(current_month, months, day)">
+ <option value='' py:attrs="{'selected': 'selected' if not current_month else None}">All months</option>
+ <py:for each="i, (val,desc) in enumerate(months)">
+ <option value='${val}' py:attrs="{'selected': 'selected' if current_month == val else None}">${desc}<py:if test="i == 0 and day"> (up to ${day})</py:if></option>
+ </py:for>
+</select>
+
<table py:def="social_table(items, with_source=False)" class="table table-condensed table-bordered table-striped">
<tr>
--- a/ckanext/ga_report/templates/ga_report/publisher/index.html
+++ b/ckanext/ga_report/templates/ga_report/publisher/index.html
@@ -27,13 +27,9 @@
<form class="form-inline" action="${h.url_for(controller='ckanext.ga_report.controller:GaDatasetReport',action='publishers')}" method="get">
<div class="controls">
- <select name="month">
- <option value='' py:attrs="{'selected': 'selected' if not c.month else None}">All months</option>
- <py:for each="val,desc in c.months">
- <option value='${val}' py:attrs="{'selected': 'selected' if c.month == val else None}">${desc}</option>
- </py:for>
- </select>
+ ${month_selector(c.month, c.months, c.day)}
+
<input class="btn button btn-primary" type='submit' value="Update"/>
</div>
</form>
@@ -41,14 +37,13 @@
<table class="table table-condensed table-bordered table-striped">
<tr>
<th>Publisher</th>
-<!-- <th>Dataset Visits</th>-->
<th>Dataset Views</th>
</tr>
<py:for each="publisher, views, visits in c.top_publishers">
<tr>
- <td>${h.link_to(publisher.title, h.url_for(controller='ckanext.ga_report.controller:GaDatasetReport', action='read_publisher', id=publisher.name))}
+ <td>
+ ${h.link_to(publisher.title, h.url_for(controller='ckanext.ga_report.controller:GaDatasetReport', action='read_publisher', id=publisher.name) + (("?month=" + c.month) if c.month else ''))}
</td>
-<!-- <td>${visits}</td> -->
<td>${views}</td>
</tr>
</py:for>
--- a/ckanext/ga_report/templates/ga_report/publisher/read.html
+++ b/ckanext/ga_report/templates/ga_report/publisher/read.html
@@ -25,12 +25,9 @@
<form class="form-inline" action="${h.url_for(controller='ckanext.ga_report.controller:GaDatasetReport',action='read')}" method="get">
<div class="controls">
- <select name="month">
- <option value='' py:attrs="{'selected': 'selected' if not c.month else None}">All months</option>
- <py:for each="val,desc in c.months">
- <option value='${val}' py:attrs="{'selected': 'selected' if c.month == val else None}">${desc}</option>
- </py:for>
- </select>
+
+ ${month_selector(c.month, c.months, c.day)}
+
<select name="publisher">
<option value='' py:attrs="{'selected': 'selected' if not c.publisher else None}">All publishers</option>
<py:for each="val,desc in c.publishers">
--- a/ckanext/ga_report/templates/ga_report/site/index.html
+++ b/ckanext/ga_report/templates/ga_report/site/index.html
@@ -24,13 +24,9 @@
<form class="form-inline" action="${h.url_for(controller='ckanext.ga_report.controller:GaReport',action='index')}" method="get">
<div class="controls">
- <select name="month">
- <option value='' py:attrs="{'selected': 'selected' if not c.month else None}">All months</option>
- <py:for each="val,desc in c.months">
- <option value='${val}' py:attrs="{'selected': 'selected' if c.month == val else None}">${desc}</option>
- </py:for>
- </select>
+ ${month_selector(c.month, c.months, c.day)}
+
<input class="btn button btn-primary" type='submit' value="Update"/>
</div>
</form>